Help!! DSP's not reporting!

(sigh)

Hi, all

I have an NM-HDV with two (2) PDVM-12s, and for some reason, the DSPs aren't showing up with any of the relevant show commands. Also, everytime I try to set up a T1 CAS link with said DSPs, the router informs me that I don't have enough DSP to complete the configuration.

% Not enough DSP resources available to configure ds0-group 0 on controller T1 1/0
% The remaining dsp resources are enough for 0 time slots.
% For current codec complexity, 1 extra dsp(s) are required to create this voice port.

?!

I get the same error in both my 3640 and my 2620XM (both with 128 MB DRAM and 12.4 IOS).

I'm befuddled - I KNOW the chips are in there, and the module doesn't physically give off any indication that it's malfunctioning (i.e., LED color). What gives? I should have six DSPs with these two chips!

...If someone could help me with this, I'd really appreciate it.

Thanks,

Volm

Find more posts tagged with

Comments

mikej412

volmeru wrote: »

I get the same error in both my 3640 and my 2620XM (both with 128 MB DRAM and 12.4 IOS).

What IOS feature set(s)?

volmeru wrote: »

the module doesn't physically give off any indication that it's malfunctioning (i.e., LED color).

Um.... does that mean you have the green LED(s) showing the DSPs are installed correctly in the 2 banks?

You can use the undocumented test dsp command mentioned in Troubleshooting the DSP on NM-HDV for Cisco 2600/3600/VG200 Series Routers - Cisco Systems to test your DSPs

volmeru

Advanced Enterprise for both (the XMs have 12.4T), yes, the LEDs are green (chips installed in banks 0 and 1), and I've tried to test the DSPs earlier today (nothing comes up; nothing is pinged).

Here's the output from earlier (salvaged from my console session):

R1#test dsp 2

Section:
1 - Query dsp resource and status
2 - Display voice port's dsp channel status
3 - Print dsp data structure info
4 - Change dsprm test Flags
5 - Modify dsp-tdm connection
6 - Disable DSP Background Status Query
7 - Enable DSP Background Status Query
8 - Enable DSP control message history
9 - Disable DSP control message history
10 - show dsp version
11 - Show alarm stats
12 - Enable dsprm alarm monitor
13 - Disable dsprm alarm monitor
q - Quit
Select option : 1

Dsp firmware version: 4.4.21
Maximum dsp count: 15
On board dsp count: 0
Jukebox available
Total dsp channels available 0
Total dsp channels allocated 0
Total dsp free channels 0
Querying dsp status......

...And that was it. Strange, huh?

Any insight into why these things aren't showing up?

Thanks in advance!

mikej412

So far based on what you posted, all your DSPs are dead on the PVDMs

I've got one dead dsp (#3) in this NM-HDV fully loaded with PVDM-12s

The bootup messages show it "missing"

*Mar  1 00:47:33.603: %DSPRM-5-UPDOWN: DSP 1 in slot 1, changed state to up
*Mar  1 00:47:33.623: %DSPRM-5-UPDOWN: DSP 2 in slot 1, changed state to up
*Mar  1 00:47:33.643: %DSPRM-5-UPDOWN: DSP 4 in slot 1, changed state to up
*Mar  1 00:47:33.663: %DSPRM-5-UPDOWN: DSP 5 in slot 1, changed state to up
....

DSP #3 is missing here too.....

R2650XM-1#test dsp 1

Section:
1 - Query dsp resource and status
2 - Display voice port's dsp channel status
3 - Print dsp data structure info
4 - Change dsprm test Flags
5 - Modify dsp-tdm connection
6 - Disable DSP Background Status Query
7 - Enable  DSP Background Status Query
8 - Enable DSP control message history
9 - Disable DSP control message history
10 - show dsp version
11 - Show alarm stats
12 - Enable dsprm alarm monitor
13 - Disable dsprm alarm monitor
q - Quit
Select option : 1

Dsp firmware version: 8.4.2
Maximum dsp count: 15
On board dsp count: 14
Jukebox available
Total dsp channels available 56
Total dsp channels allocated 0
Total dsp free channels 56
Querying dsp status......
*Mar  1 00:48:26.109: dsp 1 is ALIVE
*Mar  1 00:48:26.109: dsp 2 is ALIVE
*Mar  1 00:48:26.109: dsp 4 is ALIVE
*Mar  1 00:48:26.109: dsp 5 is ALIVE
*Mar  1 00:48:26.109: dsp 6 is ALIVE
*Mar  1 00:48:26.109: dsp 7 is ALIVE
*Mar  1 00:48:26.113: dsp 8 is ALIVE
*Mar  1 00:48:26.113: dsp 9 is ALIVE
*Mar  1 00:48:26.113: dsp 10 is ALIVE
*Mar  1 00:48:26.113: dsp 11 is ALIVE
*Mar  1 00:48:26.113: dsp 12 is ALIVE
*Mar  1 00:48:26.113: dsp 13 is ALIVE
*Mar  1 00:48:26.113: dsp 14 is ALIVE
*Mar  1 00:48:26.113: dsp 15 is ALIVE
R2650XM-1#

You can try some of the other test options.... this one shows version 0.0.0 for the "dead" #3

R2650XM-1#test dsp 1

Section:
1 - Query dsp resource and status
2 - Display voice port's dsp channel status
3 - Print dsp data structure info
4 - Change dsprm test Flags
5 - Modify dsp-tdm connection
6 - Disable DSP Background Status Query
7 - Enable  DSP Background Status Query
8 - Enable DSP control message history
9 - Disable DSP control message history
10 - show dsp version
11 - Show alarm stats
12 - Enable dsprm alarm monitor
13 - Disable dsprm alarm monitor
q - Quit
Select option : 10
dsp[0].ver_num =8.4.2
dsp[1].ver_num =8.4.2
dsp[2].ver_num =0.0.0
dsp[3].ver_num =8.4.2
dsp[4].ver_num =8.4.2
dsp[5].ver_num =8.4.2
dsp[6].ver_num =8.4.2
dsp[7].ver_num =8.4.2
dsp[8].ver_num =8.4.2
dsp[9].ver_num =8.4.2
dsp[10].ver_num =8.4.2
dsp[11].ver_num =8.4.2
dsp[12].ver_num =8.4.2
dsp[13].ver_num =8.4.2
dsp[14].ver_num =8.4.2

R2650XM-1#

The only other think I can think of is trying the PVDMs one at a time, starting with the PVDM in bank 1 (and make sure you have it seated properly in bank 0)

You could also try different IOS releases (and hope you found the same bug in the 2600/3600 IOS versions you tried).

volmeru

Trying them one at a time now...

Something funny, though - they DO show up with a 'show diag'...

mikej412

volmeru wrote: »

they DO show up with a 'show diag'

You said the LEDs showed green on the NM-HDV for Bank 0 and Bank 1, so you'd expect the PVDMs to show up in the show diag

Here's what I've got on my XM

        HDV SIMMs: Product (FRU) Number: PVDM-12=
         SIMM slot 0: PVDM-12 SIMM present.
         SIMM slot 1: PVDM-12 SIMM present.
         SIMM slot 2: PVDM-12 SIMM present.
         SIMM slot 3: PVDM-12 SIMM present.
         SIMM slot 4: PVDM-12 SIMM present.

But that just shows that the SIMMs are plugged in the slots and recognized -- not that the DSPs on the PVDMs are working.

But there is some good news -- now I don't feel so bad having one bad DSP out of all my PVDM-12s

volmeru

Mike, I think I have the answer - it's a hardware issue. Slot 0 on the NM-HDV is not working. It's dead. I only have two chips, as I said before, so I reseated them in slots 3 and 4, and they all show up with a 'test dsp 2', now.

I don't know if I should be glad or not (just a lab; I'm working on my CCVP).

Thanks a bunch for you help and lightning-quick response!